{"skill":{"slug":"agent-memory-tiers","displayName":"Agent Memory Tiers","summary":"Tiered memory system for OpenClaw agents. Gives agents instant context on startup with a 4-line state snapshot (L0) and 7-day rolling context (L1). Eliminate...","tags":{"latest":"1.0.0"},"stats":{"comments":0,"downloads":166,"installsAllTime":0,"installsCurrent":0,"stars":0,"versions":1},"createdAt":1773751291994,"updatedAt":1773751318548},"latestVersion":{"version":"1.0.0","createdAt":1773751291994,"changelog":"Initial release of agent-memory-tiers: structured, self-updating memory for OpenClaw agents.\n\n- Provides a two-tier memory system (L0 state snapshot, L1 7-day rolling context) for instant onboard context at every activation.\n- Prevents token waste by eliminating the cold-start context problem in agents.\n- Includes setup steps with example templates for L0.md and L1.md files, along with guidance for integrating into agent workflows.\n- Designed for multi-agent swarms; enables quick agent status checks and coordination via L0/L1 files.\n- Tested across a 20-agent production swarm, delivering significant token savings per activation.","license":"MIT-0"},"metadata":{"os":null,"systems":null},"owner":{"handle":"dirtyrootsstudio","userId":"s174g2kxggw2axp3btn8fdm53983gtrm","displayName":"Swarm Forge","image":"https://avatars.githubusercontent.com/u/248451410?v=4"},"moderation":null}